Oracle Cloud Infrastructure (OCI) is excited to announce that we’re supporting the Student Cluster Competition and the first ever Data Science Competition at the SC21 conference in St. Louis, Missouri, this November 14–19.
In the Student Cluster Competition, students compete with hardware that they bring to the conference with the added might of OCI’s high-performance computing (HPC) bare metal offerings, high-performance storage, and high-speed low-latency RDMA networking. Teams have access to the latest Intel Ice Lake processors through our BM.Optimized3.36 shapes and the latest NVIDIA A100 Tensor Core GPUs through our BM.GPU4.8 shapes.
In the Data Science Competition, students use Oracle’s latest Intel Ice Lake processors and NVIDA V100 GPUs and have access to the powerful Oracle data science platform, which improves productivity with unparalleled platform abilities, such as batch jobs, model catalog, and model deployment.
Students compete with hybrid cloud HPC
Teams of undergraduate students have been coming to Supercomputing (SC) to compete in the Student Cluster Competition (SCC) since 2007. This competition aims to immerse teams in the real-world challenges of running a modern hybrid cloud HPC system with obstacles such as fluctuating power caps for on-premises systems and balancing workloads across on-premises and cloud HPC systems. This year, four teams compete nonstop for 48 hours at the conference to complete real-world scientific workloads, showing off their HPC knowledge for conference attendees and judges.
Data Science takes the stage
With the explosive growth of data, organizations need to be able to extract useful features and insights by using AI and ML algorithms running on high-performance NVIDIA GPU’s and Intel CPUs. These reasons have made the field of data science quickly become one of the most important areas for organizations and HPC teams to support. For this reason, SC21 is holding their first ever Data Science Competition (DSC), where five teams across the world compete in two phases before the actual conference and present their projects to be judged at the conference. This competition runs entirely on OCI’s high-performance Intel CPUs and NVIDIA GPUs along with our data science platform.
The data science platform enables data scientists to build, train, and manage machine learning models on Oracle Cloud using an open source Python ecosystem, which includes NVIDIA RAPIDS accelerated libraries and is enhanced by the Oracle Accelerated Data Science SDK (ADS) to provide features like model evaluation, model explanation, AutoML, and more.
Science & Beyond
SC has always been the major forum where the latest ideas, technologies, and performance gains are shared with the HPC community. Having the Student Competitions at SC allows these ideas to extend toward a new generation and provides these students an opportunity to showcase their expertise and inspire them to pursue careers in the HPC field. SC’s tagline this year, “Science & Beyond” highlights the way super-computing continues to grow and be used for a broader range of efforts.
With the support of both the Student Cluster Competition and the Data Science Competition, Oracle is very excited to be a part of this discovery, which we consider to be critical for the future generation of HPC professionals, as more organizations move to embrace hybrid cloud HPC and cloud HPC to become more agile and scalable. Oracle Cloud Infrastructure is ready to take students and organizations to the next level.